Find research papers on a workplace question with Consensus

When a manager asks "does hybrid work reduce turnover?", the useful answer is what the research says, how strong it is, and whether it applies to your company. Consensus searches research papers and shows how many support, contradict or are neutral on a claim. In about 30 minutes you can go from a question to a short evidence review with references. The meter is a starting point; the conclusion comes from reading the strongest studies.

Phrase a testable question

Turn the manager's question into a statement research can test, for example "does a hybrid work format reduce staff turnover". Vague questions ("is hybrid work good?") return a mix of studies on productivity, wellbeing and commuting that do not add up to an answer. Note your company's context as well: industry, size, country. It decides later whether the evidence applies.

Search for reviews first

Search Consensus for reviews and meta-analyses on the question before individual studies. A good review has already done the weighing across many papers. Then add individual studies to fill gaps, especially recent ones. The database is in English, so search in English.

Choose 5 to 10 studies

Select 5 to 10 studies with the largest samples and recent dates. This is where the meter can mislead. It counts papers without weighing their quality, so ten small studies agreeing can look like consensus against one large, careful study that found the opposite. Sample size, study design and date matter more than the count.

Read the results

Open each paper and read the results section. Note the effect, its size and the conditions: which industry, which country, which kind of employees. AI summaries simplify results, often dropping the conditions under which an effect holds. Take numbers from the paper itself. Then check whether later work confirms the conclusions.

To structure the review, paste the abstracts and results into a chat assistant with this prompt:

Role: you are a researcher who prepares short evidence reviews for managers. Context: question: {{question}}. Company context: {{industry, size, country}}. I am adding the abstracts and results of the studies found. Task: summarize the evidence: what the studies say, the effect size, the samples and countries, where results contradict each other. Rate the strength of evidence: strong, moderate, weak. Format: a two-sentence conclusion, a table "study, year, sample, result", and caveats on applying it in {{country}}. No long dashes.

Write the conclusion honestly

State the conclusion with its strength of evidence and the conditions. "Moderate evidence that hybrid work reduces turnover among office staff in large companies; little evidence for small firms" is useful. "Research shows hybrid work reduces turnover" overstates it. One striking study with a small sample is easy to pass off as consensus; avoid doing that in your own review.

What to do

1. Phrase the question as a testable statement and note your company's context.
2. Search Consensus for reviews and meta-analyses first.
3. Select 5 to 10 studies with the largest samples and recent dates.
4. Read the results sections and note effect, size and conditions.
5. Check whether later work confirms the conclusions.
6. Write the conclusion with the strength of evidence and references.

A sound evidence review

The conclusion rests on several studies with references, and it names the strength of evidence and the conditions under which the effect holds.

Pricing

Consensus offers unlimited search and a limited number of AI summary queries a month for free. Premium costs 8.99 USD per month with annual billing (11.99 monthly) and adds unlimited summaries, deep search and export; Teams is 9.99 USD per user with annual billing. See Consensus pricing, the recruiter's version of this task and Consensus vs Scite.